planet10 said:
that is only half true: don't need to show that for the source material to be an issue, you have to prove that it is "confunding" the results?
all we have here are factors under which the source material "could" confunding the results.
The question's backward: how do you know what people can hear absent any rigorously controlled test?
Please explain why ABX is "statistically unable" to test whether or not switching a specific component in a sound system is humanly audible.planet10 said:
You have not made a coherent argument against the protocol; if it exists elsewhere, in this thread or elsewhere on the internet, I'd like to read it.
I don't think anybody has ever claimed that ABX testing "captures all audio distinctions".serengetiplains said:
It's a method of doing a double-blinded, randomized test of the hypothesis that two specific audio components are not audibly distinguishable.
Or were you making a funny?
You would have to find a new sample of listeners and repeat the test.planet10 said:
Or use the same listeners, but on a different day and with different test equipment.
No different than any other psychological test.
If ABX doesn't capture all audio distinctions, given a negative ABX result (ie, no ABX-audible difference), how can you then say two components are producing the audibly-same sound?
You can't. Proving a negative is, as you well know, not philosophically possible. It's up to someone asserting that A is different than B to demonstrate a positive result in a controlled test (whether ABX or other).
There could be, for example, measureable differences between components which test subjects could not distinguish. That's the very essence of the ABX test, determining which differences are indeed audible to human beings.serengetiplains said:
As far as "captur[ing] all audio distinctions", I am not sure to what you are referring. I am certain we could come up with "audio distinctions" that would be all-but-impossible to test via ABX. For example, it would be very difficult to successfully blind a test of room treatments.
You are veering into philosophical territory. I mean, maybe the value of ABX testing is negative, because people's feelings get hurt when their preconceptions are challenged?serengetiplains said:
Please explain.
To take this in another direction:
Many of you may be pleased to know that I have applied to JREF to prove that an audible difference can exist in speaker cables. Perhaps even to demonstrate that one cable is "better" or at least of "higher fidelity" than another.
I'm actually skeptical that there will be enough proof to win $1,000,000. However, to rebut Fokker, I am NOT afraid of the money, nor am I afraid of the challenge.
I will not predict the outcome of the test, because I really don't know. There is nothing to lose, except some time and effort. And even that will not be a loss, if the test really goes to show there is NO difference.
Until the challenge is accepted or rejected by the JREF, I will NOT be describing it here. I'm not silly enough to have one of you guys scoop my million bucks! 😀
Just let me say that the test has nothing to do with what you have been discussing for the past dozen pages. The test is NOT subjective - no double blind needed (tho it could be used). The test is based on a well known, often used electrical test that has existed for more than 100 years.
Once my application has been either accepted or rejected, I will detail the test here.
Many of you may be pleased to know that I have applied to JREF to prove that an audible difference can exist in speaker cables. Perhaps even to demonstrate that one cable is "better" or at least of "higher fidelity" than another.
I'm actually skeptical that there will be enough proof to win $1,000,000. However, to rebut Fokker, I am NOT afraid of the money, nor am I afraid of the challenge.
I will not predict the outcome of the test, because I really don't know. There is nothing to lose, except some time and effort. And even that will not be a loss, if the test really goes to show there is NO difference.
Until the challenge is accepted or rejected by the JREF, I will NOT be describing it here. I'm not silly enough to have one of you guys scoop my million bucks! 😀
Just let me say that the test has nothing to do with what you have been discussing for the past dozen pages. The test is NOT subjective - no double blind needed (tho it could be used). The test is based on a well known, often used electrical test that has existed for more than 100 years.
Once my application has been either accepted or rejected, I will detail the test here.
The other thing that ABX has no truck with is the comparison of components that are indeed audibly different.serengetiplains said:
Such a test would be possible for certain components, you could simply blind that A and B switch and let the listener say which component he preferred. Other components might be more difficult, for example, the old "CD vs vinyl" debate would be very difficult, because the audible artifacts of each medium would easily unblind the tests.
Dumbass said:
Buried somewhere in the HiFidelity mail list archives, late 1999, or 2000. David Klien & Tim Bailey.
dave
I'm a very simple man. I do not believe in psychoacoustics, metaphysics, Socrates, Descartes or the Easter Bunny. Life is simple. Even a child can do it.
If I do an ABX test and scores 100 pct, how is that test not reliable? If I can tell a Toyota from a Mazda every single time and duplicate it at will, how is that (or me, my eyes, my memory, whatever) not reliable? The only difference here is that I can tell the difference between a Toyota and a Mazda. I cannot tell the difference between two cables. The only conclusion I can draw from that is that the shortcoming of the ABX test is the shortcoming of the cable fans.
I cannot see any difference between this reasoning and the phoney Stereophile article. One bends the truth to fit his beliefs, the other rules out a test because he doesn't like the result.
To corrupt SY's post. The ABX test is not designed to find out what you cannot hear. It's designed to find out what you can hear. There are many reasons for why you might fail a test. But there's only one reason for why you passed it--you could tell one cable from another.
If I do an ABX test and scores 100 pct, how is that test not reliable? If I can tell a Toyota from a Mazda every single time and duplicate it at will, how is that (or me, my eyes, my memory, whatever) not reliable? The only difference here is that I can tell the difference between a Toyota and a Mazda. I cannot tell the difference between two cables. The only conclusion I can draw from that is that the shortcoming of the ABX test is the shortcoming of the cable fans.
I cannot see any difference between this reasoning and the phoney Stereophile article. One bends the truth to fit his beliefs, the other rules out a test because he doesn't like the result.
To corrupt SY's post. The ABX test is not designed to find out what you cannot hear. It's designed to find out what you can hear. There are many reasons for why you might fail a test. But there's only one reason for why you passed it--you could tell one cable from another.
Doubt that this will be acceptable; it's not paranormal. Hey, with an impedance bridge and/or an RF generator, I can easily distinguish between zip cord and, say, MIT or Nordost. That's different than, absent linear errors, being able to audibly distinguish them.
Dumbass said:
Care to answer by stating how much of the audible universe exists within ABX testing?
- Status
- Not open for further replies.